An Overview of the Damaging and Low

Bulletin of the Seismological Society of America, Vol. 97, No. 3, pp. 671–690, June 2007, doi: 10.1785/0120050150

An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005: Context, Seismotectonics, and Seismic Risk Implications for Southeast Spain by B. Benito, R. Capote, P. Murphy, J. M. Gaspar-Escribano, J. J. Martıńez-Dıáz, M. Tsige, D. Stich, J. Garcıá-Mayordomo, M. J. Garcıá Rodrı´guez, M. E. Jimeńez, J. M. Insua-Are´valo, J. A. A´ lvarez-Go´mez, and C. Canora

Abstract This article presents an overview of the La Paca earthquake of magnitude mbLg 4.7, which occurred on 29 January 2005, with its epicenter located near the town of Avile´s in the Murcia region in southeast Spain. Despite its low magnitude, the earthquake caused important damage in two towns of the epicentral area, La Paca and Zarcilla de Ramos. These areas recorded intensities of VI–VII (European Macroseismic Scale, 1998) and sustained estimated economic losses amounting to 10 million €. Aftershocks continued for more than 2 weeks, producing considerable alarm in the population and mobilizing emergency services from the whole region. The La Paca seismic series is the third registered in the region in the past 8 years, being preceded by the Mula (1999) and southwest Bullas (2002) seismic series. These

main events had also low magnitudes (mbLg 4.8) and caused damage levels similar to the 2005 earthquake. The case is an example of a moderate seismic zone where low-magnitude and frequent earthquakes have important implications on the seismic hazard and risk of the region. Although these are not the largest expected earthquakes, they have yielded important information for improving the knowledge of the seismic activity of the area. With this aim in mind, different topics have been analyzed from a multidisciplinary perspective, including seismicity, local tectonics and surface geology, focal mechanisms, macroseismic effects, and ground motion. Results indicate a local tectonic interpretation, consistent with a strike-slip focal mechanism, the confirmation of a triggering process between the 2002 and 2005 earthquakes, a geotechnical and ground-motion characterization for the damaged sites (supporting local amplification effects and estimated peak ground acceleration values of 0.1g), and an understanding of damage patterns in relation to local building trends. The results may be used as guidelines for future revisions of the Spanish Building Code (Norma de la Construccioń Sismorresistente Espanõla [NCSE-02], 2002). The study results should contribute to risk mitigation in a region where strong-motion records from the maximum expected earthquakes are not available. This approach can be extended to other regions with similar seismic backgrounds and a lack of strong-motion records.

Introduction The 2005 series began on the 29 January, when an earth- [EMS-98]), causing important economic losses but fortu- quake of magnitude mbLg 4.7 or Mw 4.8 took place near nately no casualties. We refer to this earthquake as the 2005 the town of Avile´s (Murcia region, southeast Spain). This is La Paca earthquake in allusion to one of the most damaged the third earthquake resulting in considerable damage in the towns. Murcia region in the past 8 years, following the earthquakes The Murcia region is an area with moderate seismic of Mula (2 February 1999, mbLg 4.8) and southwest Bullas activity in Spain, with a regional stress ﬁeld related to the (6 August 2002, mbLg 4.8). The three earthquakes reached convergence of the African and Eurasian plates. In this area, intensities of VI–VII (European Macroseismic Scale 1998 relatively frequent earthquakes with low-to-moderate mag-

671 672 B. Benito et al. nitudes take place, having caused damage in historical and resistant building Code NCSE-02 (Norma de la Construccioń recent times. Temporal clustering of significant events has Sismorresistente Espanõla [NCSE-02], 2002), shown in eventually been identified in the region. Since 1999, three Figure 1. The represented value is the NCSE-02 basic accel- damaging earthquakes have occurred in a short time. This eration, ab, corresponding to horizontal peak ground accel- timing raises the possibility of a triggering mechanism that eration (PGA) in hard soil with a 10% probability of being may initiate further events in the near future, provoking cor- exceeded in 50 years. Expected ab values for the Murcia responding concerns for the regional seismic risk and hazard region range between 0.04 and 0.16g. Alongside Granada status. and Alicante, Murcia is one of the regions with the highest The heaviest damage in the 2005 earthquake was not hazard levels in Spain. observed in the town nearest to the epicenter (Avile´s), but in other more distant locations (La Paca and Zarcilla de Ramos), where first assessments established macroseismic Regional Tectonic Setting intensities of VI and VII (EMS-98), respectively. Several The study area is located in the external zone of the factors influencing the damage distribution are analyzed: site Southeastern Betic Cordilleras (Fig. 2a). The epicentral re- conditions, input ground motion, and the vulnerability of gion is characterized by a complex tectonic structure due to local construction trends. its position between a major shear zone, the Crevillente Fault Ground-motion values of up to 0.024g were obtained Zone, also known as the Ca´diz-Alicante Fault (Sanz de Gal- for the 2005 series at strong-motion stations located more deano 1983), and the contact between the Internal and Ex- than 20 km from the epicenter of the main event. No strong- ternal zones (Hermes 1978, 1985). The Crevillente fault is motion records are available in the epicentral area for direct a 150-km-long dextral strike-slip fault, active during the late characterization of the ground motion in that location. This Miocene and early Tortonian (Martıńez-Dıáz, 1998; Sanz de same problem was encountered for the 1999 and 2002 earth- Galdeano and Buforn, 2005). Strike-slip motion along the quakes. For this reason, the simulation of the ground motion fault has contributed to the westward emplacement of the is of particular interest for inferring strong-motion and re- Alborań Domain units of the internal zones (Andrieux et al., sponse spectra values for both the epicentral area and the 1971; Sanz de Galdeano, 1990). These movements produced locations of the most damaged towns of La Paca and Zarcilla a regional morphostructure characterized by northeast– de Ramos. In these locations damage was reported to modern southwest-trending ranges of Jurassic limestones separated code-compliant engineered structures. This is a critical point by Triassic gypsums and Cretaceous-Tertiary marls. regarding revisions of the building code and future risk mit- Since the Late Miocene, the Crevillente fault has been igation in the region. unfavorably aligned for strike-slip motion because its N70E This article presents an overview of the conclusions re- trend is normal to the current regional shortening direction sulting from a multidisciplinary analysis of this striking case, (Sanz de Galdeano and Buforn, 2005). During the Upper incorporating geological, seismological, and architectural Miocene-Pliocene, marine basins were developed to the data. The purpose is twofold: (1) to gather as much infor- north of the Crevillente fault, while to the south, Pliocene mation as possible regarding the earthquake’s characteristics lakes filled the depressions between the Sierras. N50E and effects, such as the spatial and chronological distribution strike-slip left-lateral faults, coherent with the current stress of events, focal parameters of significant shocks, strong- field, have been created or reactivated in the region since the motion records, and damage reports, and (2) to infer from Tortonian onward, affecting the Pliocene and the Pleistocene these data explanations for the unexpected damage distri- deposits. Paleoseismic studies on some of these faults (i.e., bution, evidence relating this event back to the 2002 south- the Alhama de Murcia Fault; Fig. 2b) report a significant west Bullas event, and possible implications for the regional palaeoseimic activity during the Holocene and several sur- seismic-hazard levels. Achieving this purpose requires the face ruptures have been identified (Martıńez Dıáz et al., analysis of local-scale fault kinematics, local soil conditions, 2001). and local building practices.

Historical and Instrumental Seismicity in the Region Seismic Hazard and Regional Seismotectonic The historical seismicity reported for the Murcia region Framework for the Murcia Region includes several damaging earthquakes. The location and intensity of these earthquakes, with estimated intensities I Ն Seismic Hazard VII are given in Table 1. Epicenters are shown in Figure 3. The epicenter of the 2005 La Paca earthquake, as well The first earthquake of which we have knowledge oc- as those of the previous series of 1999 and 2002, is located curred in 1579, with intensity VII. Another earthquake in the southwest of the province of Murcia, a zone with reached intensity VIII in 1674, and two additional earth- moderate seismic hazard. A first view of the relative hazard quakes are reported in 1911 with the same intensity, causing for this zone compared with other regions in Spain is ob- significant damage in Torres de Cotillas and Lorquı´. These tained from the hazard map of the Spanish earthquake- earthquakes have been the subject of several reports: An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005 673

Figure 1. Seismic-hazard map of the Spanish Building Code NCSE-02. Isolines represent “basic acceleration, ab,” which corresponds to the expected PGA in hard soil for exceedence probabilities of 10% in 50 years. The Murcia province is framed with a rectangle.

Figure 2. (a) Geological map of the Betic Cordillera and surrounding areas. The rectangle shows the location of the studied area. CRF, Crevillente fault; AFZ, Alpujarras fault zone; CFZ, Carboneras fault zone; PF, Palomares fault. (Modified from Martıńez-Dıáz, 1998.) (b) Principal faults with Quaternary activity identified in the area around the epicenter of the 2005 Bullas sequence. CRF, Crevillente fault; AMF, Alhama de Murcia fault; PF, Palomares fault; BSF, Bajo Segura fault. (Modified from Martıńez- Dıáz, 1998.) 674 B. Benito et al.

Table 1

Location Data, Epicentral Intensity (I0, EMS) and Magnitude (mbLg) of the Earthquakes That Have Occurred in Murcia Ն Province with I0 VII and/or mbLg 4.0

Date Time UTC Longitude Latitude Depth Maximum Intensity Magnitude (yyyy/mm/dd) (h:m:s) ( W) ( N) (km) Location (EMS) mbLg 1579/01/30 ? 1.70 37.68 Lorca VII 1674/08/28 21:30:00 1.70 37.68 Lorca VIII 1743/03/09 16:00:00 1.13 38.00 Murcia VII 1907/04/16 17:30:00 1.50 37.80 Totana VII 1908/09/29 00:00:00 1.30 38.10 Ojo´s VII 1911/03/21 14:15:35 1.21 38.01 Torres de Cotillas VIII 1911/04/03 11:11:11 1.20 38.10 Lorquı´ VIII 1917/01/28 22:32:31 1.26 38.03 Torres de Cotillas VII 1930/09/03 09:59:58 1.23 38.06 Lorquı´ VII 1944/02/23 22:34:10 1.15 38.16 Fortuna VII 1948/06/23 03:43:55 1.75 38.14 Cehegı´n VIII 1972/04/14 03:22:17 1.35 38.47 5 Jumilla 4.2 1977/06/06 10:49:12 1.73 37.64 9 Lorca VI 4.2 1978/03/24 13:01:24 1.70 37.63 5 Lorca 4.3 1995/11/26 05:39:40 1.27 38.03 2 N. Alcantarilla V–VI 4.1 1996/09/02 19:07:01 1.55 37.56 1 Mazarro´n V 4.5 1999/02/02 13:45:18 1.50 38.09 1 Mula VI 4.8 2002/08/06 06:16:19 1.83 37.88 1 Southwest Bullas VI* 4.8 2005/01/29 07:41:31 1.78 37.88 3 La Paca VII* 4.7

Data extracted from the IGN database (Me´zcua and Martı´nez Solares, 1983; Martı´nez Solares and Me´zcua, 2002; and the IGN website, www.ign.es) except (*), which are our own data.

Figure 3. Epicenters of the major earthquakes that have occurred in Murcia Prov- Ն ince (I0 VII and/or mbLg 4.0). Location parameters are given in Table 1. An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005 675

Jimeńez Cisneros (1911), Sańchez Lozano and Marıń Seismic Description of the 2005 Series: Spatial (1912), Sańchez Navarro-Neumann (1911, 1912, 1917, and Chronological Distribution 1920), Galbis (1932), and Rey Pastor (1936). The magni- The La Paca earthquake was located near the town of ע ע tudes were estimated as Ms 5.7 0.5 and 5.3 0.3 for the Avile´s (Fig. 4). This shock was followed by an important March and April 1911 events, respectively (Buforn et al., aftershock series lasting more than 2 weeks. The epicenters 2005). of the seismic series are shown in Figure 4, together with After the 1911 series, no sizable earthquakes occurred epicenters of the previous series of 1999 and 2002 (locations in the Murcia region, until an event on 23 June 1948, located given by IGN). As we can see from the figure, the epicenters 60 km to the west of the 1911 series (Fig. 3). According to of the 2002 and 2005 series partly overlap each other. Rey Pastor (1949), the maximum intensity of the 1948 event The epicenters of the events following the mainshock was VIII Forell–Mercalli and its magnitude was 5.3 (Badal between 29 January and 7 February are represented in 2-day et al., 2000). intervals in Figure 5. Analysis of these events suggests that The seismic activity of the area has continued during the 29 January shock triggered an aftershock sequence with the period of instrumental records. Table 1 includes the data a north–south to northeast–southwest trend. The activity of all the earthquakes from 1972 onward with magnitude started to decrease after 31 January. On 3 February a new greater than or equal to 4.0. The epicenters of these earth- event with magnitude 4.2 (mbLg) took place, triggering an- quakes are shown in Figure 3 together with the epicenters other aftershock series, extending predominately northward of historical earthquakes. from the epicenter. The aftershock activity was thereby re- Between the last historical event of 1948 and 1996, five activated by this event. Figure 6a shows a representation of earthquakes took place with magnitudes between 4.0 and 4.5 the number of events per day. The number of events clearly (mbLg). On 2 February 1999, an earthquake with mbLg 4.8 decreased during the 2 days after the mainshock of occurred near the town of Mula (Fig. 3), starting a period of 29 January, rising again after the 3 February event. The higher activity in the region. A maximum intensity of VI whole process seems to indicate that the two ruptures took (EMS) was observed in Mula, Torres de Cotillas, Campos place on different faults or fault segments. del Rıó, and other towns along the Mula river. This earth- We can support this hypothesis by comparing the decay quake was felt at far distances, such as Madrid (400 km). of aftershock activity with the common exponential pattern This event was preceded by a foreshock and followed by proposed by Omori (1884), which in logarithmic form cor- b • log t, with מ a ס (numerous aftershocks, two of them with magnitudes greater responds to the expression: log N(t than 3.5. The mainshock has been the subject of several N(t) being the number of events per day and t the time (in studies (Instituto Geogra´fico National [IGN], 1999; Buforn days) elapsed after the mainshock. A fit to the complete se- and Sanz de Galdeano, 2001; Martıńez-Dıáz et al., 2002; ries yields a poor correlation coefficient, whereas indepen- Mancilla et al., 2002). dent fits for both seismic subseries confirm that these adjust Another earthquake with mbLg 4.8 (IGN) occurred on better to Omori’s law. Figure 6b (left and right) shows the 6 August 2002 near the town of Bullas, 20 km west of Mula. fit for the two subseries. The so-called southwest Bullas earthquake reached a maxi- Finally, we have studied the magnitude–frequency dis- mum EMS intensity of VI in Bullas and Cehegıń and was tribution of the aftershocks, comparing it with the common ס (also followed by numerous aftershocks, three of them ex- power law of the Gutenberg–Richter relation (log N(m -bm, where m is the magnitude and N(m) is the cu מ ceeding magnitude 3.5. a Based on the analysis of historical and instrumental mulative number of earthquakes with magnitude M Ն m). for the first, and 0.64 ס b ,2.71 ס events, we observe an activity pattern composed of a cluster We found a value of a -for the second subseries. Large corre 0.90 ס b ,3.56 ס of moderate magnitude events (M 5) lasting a few years, a ,respectively ,0.99 ס and R2 0.96 ס separated by longer periods of quiescence. This pattern can lation coefficients of R2 be observed between the years of 1907 and 1917 and the indicate that both subseries follow closely a magnitude– ס 1940s, with maximum intensities of I VIII (see Table 1). frequency distribution consistent with the Gutenberg– Richter law.

Local Seismotectonic Setting for the La Paca Focal Mechanism (2005) Earthquake The 2005 La Paca earthquake sequence was recorded at In this section we analyze the geographical and chro- a dense regional network of broadband seismographs. We nological distribution of the 2005 seismic series in terms of use waveforms from 12 near-regional stations to estimate the tectonic setting of the epicentral area. The 1999 and 2002 moment-tensor mechanisms and magnitudes for the La Paca series are also analyzed here because of their relevance to mainshock and major aftershocks. We represent seismic the 2005 series on account of their geographical and chro- waveforms as a linear combination of 1D Greens’ functions nological closeness. for an average, regional velocity, and density model (Stich 676 B. Benito et al.

Figure 4. Epicenters of the mainshocks and aftershocks for the 1999 Mula, 2002 Bullas, and 2005 La Paca series. Main faults are also shown in the map. et al., 2005), and invert for the deviatoric moment tensor by waveforms (Fig. 7). The aftershock mechanisms show larger minimizing the L2-misfit between observed and predicted normal components than the mainshock, specifically rather time-domain displacement seismograms (Langston et al., steep east-southeast–west-northwest striking right-lateral 1982). Waveforms are filtered in an intermediate period nodal planes, and quite shallow north–south striking left- band (20–50 sec for the mainshock) to obtain appropriate lateral nodal planes. This suggests considerable geometric path corrections. A more detailed discussion of the inversion complexity and an off-fault origin for these aftershocks. This scheme is given in Stich et al. (2003). result agrees with the hypothesis of two possible ruptures, Moment-tensor inversion for the 2005 La Paca as deduced from the temporal pattern of aftershock activity mainshock yields a seismic moment of 1.6•1016 N m and a (see “Seismic Description of 2005”). The earthquakes of moment magnitude Mw 4.8. The best moment tensor corre- 3 and 4 February show nearly identical broadband wave- sponds to a nearly double-couple force system (3% non- forms and can be considered a duplet indicating repeated double-couple reminder [CLVD]), showing predominately rupture of the same fault patch (Geller and Mueller, 1980). strike-slip movement with a minor normal component (Ta- This is consistent with our nearly identical moment-tensor bles 2 and 3). Our result is similar to an automatic near-real- mechanisms for these events. time moment-tensor estimate by IGN-Madrid (www.ign.es). For the 1999 and 2002 sequences, several focal mech- The right-lateral nodal plane has strike N132E, dip 85, and anisms have been given by routine moment tensor projects ,(the left-lateral nodal plane has strike N40E, at Istituto Nazionale di Geofisica e Vulcanologia (INGV ,153מ rake This solution corresponds to gener- Italy (Pondrelli et al., 2002), ETH Zu¨rich, Switzerland .5מ dip 63, and rake ally good waveform matches, except for unmodeled near- (Braunmiller et al., 2002), and Instituto Andaluz de Geofı´s- field contributions (Fig. 7). ica (IAG)-Granada, Spain (Stich et al., 2003), as well as in In the same way, we obtain stable inversion results and individual studies. For the 2002 series, we reprocessed mo- adequate waveform matches for three events of the La Paca ment-tensor estimates from the IAG catalog, including fur- aftershock sequence, with moment magnitudes between 3.6 ther near-regional recordings from the temporary TEDESE and 4.2 (Table 2). For all events, we have estimated the network (ROA/UCM/Geofon). Moment-tensor solutions for upper crustal source depths. Moment-tensor mechanisms the mainshock (Mw 5.0) and three aftershocks (Mw 3.5–4.0, differ noticeably from the mainshock mechanism, supported www.ugr.es/iag/tensor) show oblique normal faulting, by the corresponding differences in intermediate period with north–south (N0EtoN11E) and west-northwest– An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005 677

Figure 5. Space–time distribution of epicenters of the 2005 seismic series. The epicenters of the aftershocks are shown in time windows of 2-day duration.

east-southeast (N115E to N130E) striking nodal planes, Local Tectonic Setting very similar to the M 3.6 aftershock of 1 February 2005. w Structure in the Epicentral Area. To correlate the recent For the 1999 Mula earthquake, two similar moment- seismic activity in the La Paca area with the main tectonic tensor estimates are available from full-waveform inversion structures in the region, we have carried out a geological at seven and eight near-regional stations, respectively field survey and a review of the existing geological 1:50,000 (Mancilla et al., 2002; Stich et al., 2003), indicating mag- scale maps (Kampshuur et al., 1972a,b; Jerez Mir et al., nitude Mw 4.7–4.8, and predominately strike-slip faulting, very similar to the 2005 La Paca mainshock. In contrast, the 1974; Velando and Paquet, 1974). This study was comple- focal mechanism from first motion polarities (Buforn and mented with the analysis of field evidence for possible Qua- Sanz de Galdeano, 2001) indicates predominately reverse ternary deformation with the aim of characterizing the pos- faulting, in good agreement with the INGV moment tensor sible source of the earthquakes. We found several high-angle from the long-period, sparse network data. We prefer the brittle postalpine or neotectonic faults (Fig. 8), which are strike-slip solution (Table 2), because the first-motion po- good candidates for the source of the La Paca seismic series. larity pattern is equally consistent with both a strike-slip and The main characteristics of these faults are summarized in a reverse solution (Buforn et al., 2005), and the incorpora- Table 4. Most of these faults show a curved geometry in the tion of intermediate-period surface wave observations from epicentral area (Avile´s-La Paca-Zarcilla de Ramos), indi- the Spanish network clarifies this ambiguity (in this case cating bending of the mainfault traces. increasing L2 misfit by a factor of 2 for the reverse solution; The most relevant fault group is in a zone that crosses see Mancilla et al., 2002). the epicentral area from northeast to southwest and probably 678 B. Benito et al.

Figure 6. (a) Number of events per day, from 29 January until 18 February 2005. Two different series may be observed. (b) Adjustment to the Omori law for the two seismic series: from 29 January until 2 February (left) and from 3 February until ,for the ﬁrst 1.63 ס b ,1.82 ס February 2005 (right). The inferred parameters are a 28 for the second subseries. The corresponding correlation 2.08 ס b ,3.38 ס and a .respectively ,0.74 ס and R2 0.80 ס coefﬁcients are R2

Table 2 Location Parameters, Magnitude, and Moment Tensor Solutions for the 2005 La Paca and the 2002 Bullas Sequence (This Study), and the 1999 Mula Mainshock (Stich et al., 2003)

Date Latitude Longitude Depth Fault-Angle Parameters CLVD (yyyy/mm/dd) Time ( ) ( ) (km) (Strike/Dip/Rake) (%) M0 (N m) Mw 4.8 1016•1.65 8 25מ/41/69 ;157מ/141/66 6 1.49מ 38.11 13:45:17 1999/02/02 5.0 1016•3.47 3 27מ/11/53 ;140מ/119/68 8 1.82מ 37.91 06:16:19 2002/08/06 3.9 1014•8.81 15 28מ/0/35 ;122מ/115/73 6 1.81מ 37.89 11:55:16 2002/08/06 3.5 1014•2.34 13 27מ/7/37 ;124מ/120/74 6 1.83מ 37.86 00:43:56 2002/08/07 4.0 1015•1.08 15 33מ/11/37 ;122מ/130/70 6 1.84מ 37.86 23:09:07 2002/08/07 4.8 1016•1.62 3 5מ/40/63 ;153מ/132/85 10 1.78מ 37.88 07:41:31 2005/01/29 3.6 1014•2.61 6 35מ/1/40 ;125מ/120/68 6 1.70מ 38.02 23:53:56 2005/02/01 4.2 1015•2.44 6 7מ/15/46 ;136מ/110/84 10 1.79מ 37.82 11:40:33 2005/02/03 3.9 1014•8.65 3 10מ/11/47 ;136מ/109/82 6 1.80מ 37.82 01:09:41 2005/02/04 Fault angle parameters for both nodal planes represent the double-couple component. The non-double-couple reminder (CLVD) of the moment tensor is given in percentage. M0 is seismic moment and Mw is moment magnitude. Epicenter locations and origin times are taken from the IGN bulletin (IGN data ﬁle, www.ign.es). continues out beyond the studied zone. The structures form- show a restraining stepover in the La Paca area with right- ing this system are the Avile´s, Don Gonzalo, and Zarcilla stepping and overlapping structural geometry. The 2.5-km faults (Fig. 8), all of them with left-lateral strike-slip activity overlap and 2-km separation among these faults produces and exhibiting similar branching and linkage patterns be- high compressive deformation and tectonic stress concen- tween fault sections. The Avile´s and Don Gonzalo faults tration. An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005 679

Table 3 Input Parameters for the Coulomb Stress Transfer Modeling for the 2002 and 2005 Mainshocks

Center of Rupture

Fault-Plane Parameters Latitude Longitude Depth Rupture Area ן ן Date (yyyy/mm/dd) (Strike/Dip/Rake) ( N) ( W) Mw (km) Length Width (km km) 2.4 ן 2.4 5 4.8 1.78 37.88 5מ/40/63 2005/01/29 3 ן 3 1 5.0 1.82 37.91 27מ/Plane A) 11/53) 2002/08/06 3 ן 3 1 5.0 1.82 37.91 140מ/Plane B) 119/68) 2002/08/06 length) is estimated using the equations of Wells and Coppersmith (1994). Note that for the 2002 event, two fault-plane solutions ן Rupture area (width are considered.

Seismotectonic Interpretations for the 2005 Main Earth- stress concentration in this relatively small area, promoting quake. Seismic data and field evidence allow us to identify periodic slip events along different faults or fault segments. the Avile´s fault as the source of the main 29 January 2005 earthquake. The calculated focal mechanism solution indicates that the active fault responsible for this shock is a N45E fault with strike-slip movement located near the town Coulomb Failure Stress Modeling of Avile´s. An alternative interpretation is that the same focal Two seismic sequences in neighboring areas with sig- mechanism is produced by a dextral strike-slip fault trending nificant material damage in the short time span of 3 years northwest–southeast, which can be related with the El Puerto justifies looking into changes in the state of stress as a pos- fault. sible cause–effect process in operation. Such changes may The geological and geomorphological field evidence advance or delay the failure of faults in the region, as pro- collected in the area around the epicenter suggests that the posed in other seismically active regions (Stein, 1999; Freed Avile´s fault is the most likely source of the 2005 mainshock. and Lin, 2001; Zeng, 2001; Pollitz, 2002). This fault was active during the Quaternary and recent geo- It is known that the stress drop on a fault plane due to logical times. Toward the southwest, at Umbrıá hill, the fault the occurrence of an earthquake increases the effective shear trace crosses and deforms a conglomerate continental stress around the rupture area (Chinnery, 1963). During the formation dated as Plio-Quaternary according to geomor- past 10 years, observations from different seismic sources phological criteria. The conglomerates show a dense set of and their magnitudes have indicated that small variations in fractures aligned N80E over the fault trace. These fractures static stress, even lower than 1 bar, are able to induce the are absent in other areas of the formation; we therefore con- reactivation of nearby faults that are close to failure. This clude that this brittle deformation is evidence of fault move- phenomenon has been described as a triggering process ment occurring after the formation of these units. (Jaume and Sikes, 1992; King et al., 1994; Harris et al., 1995; Toda et al., 1998). The triggering process may involve not The Local Tectonic Framework of 2002 Bullas Earthquake. only the generation of aftershocks or major shocks, but also

The main Bullas earthquake on 6 August 2002 (Mw 5.0 or changes of seismicity rate in a certain zone, increasing or mbLg 4.8) and its bigger aftershocks are located in an area decreasing for several months after a mainshock (Stein, 1999). about 2 to 3 km north of La Paca and away from the Avile´s The triggering effect is attributed to changes in Cou- מ מ ס fault trace. The focal mechanisms indicate normal slip (with lomb failure stress (CFS): CFS sb l (rb p), where strike-slip component) on northwest–southeast-oriented sb is the shear stress over the fault plane, rb is the normal faults. The El Puerto fault (Fig. 8), which crosses through stress, p is the fluid pressure, and l is the frictional the epicenter area, is a likely source candidate for the main coefficient. earthquake of this series. Its orientation varies from N125E Following this approach, we have estimated the stress to N100E, similar to the orientation of the planes of focal change produced by the 2002 and 2005 earthquakes by fol- mechanisms. The normal slip indicated by the focal mech- lowing the Okada (1992) method, taking a value of as the shear module and a value of 0.25 as 2מanism results is in agreement with the northwest–southeast 3.2•1010 Nm horizontal shortening direction. Poisson’s coefficient. The apparent friction coefficient is Two theories may be proposed to explain why the series taken as 0.75, which is an acceptable value as proposed by of 2002 and 2005 have occurred in the same region within several authors from the study of 10 years of activity in a short time span. The first one would be a triggering process southern California (Toda and Stein, 2003). The orientation initiated after the 2002 series, caused by static stress transfer of the planes was selected from the moment-tensor mecha- produced by the El Puerto fault motion. The other one would nisms discussed previously (Table 3). The ruptures are mod- be related to the local tectonic configuration of the La Paca- eled as rectangles with a rupture area estimated by using the Avile´s area, where the restraining stepover between the equations of Wells and Coppersmith (1994). The parameters Avile´s and Zarcilla faults (Fig. 8) may produce continuous of the models are described in Table 3. 680 B. Benito et al.

Figure 7. Sample waveform matches for moment-tensor solutions of the 2005 La Paca mainshock and the 4 February aftershock. Each waveform panel shows, from top to bottom, radial, transverse, and vertical intermediate period displacement waveforms (solid lines, observed seismograms; dotted lines, moment tensor predictions). The map shows the distribution of regional broadband stations used, and moment-tensor estimates for the mainshock and largest aftershocks (Tables 2 and 3). The remaining solutions are posted online at the IAG-Granada moment tensor project (www.ugr.es/iag/tensor).

Figure 9 shows the stress transfer produced by the 2002 the 2005 event (Fig. 9a, b). The stress change produced by and 2005 mainshocks considering the two fault plane solu- the two mainshocks is calculated on optimally oriented tions for the 2002 mainshock. The stress changes produced planes for a strike-slip stress ﬁeld with a maximum horizon- by this event are calculated on planes parallel to planes ori- tal stress 169 N, in accordance with the present stress de- ented 40 N dipping 63 to the southeast, which is the source terminations of Stich et al. (2006) (Fig. 9c, d). These models plane supported by geological data to be the responsible of show that the 2005 event occurred in an area where the An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005 681

Figure 8. Simpliﬁed geological and tectonic map of the epicentral area, showing the main faults and focal mechanisms solutions for the 2002 and 2005 mainshocks. AF, Avile´s fault; TF, Terreras fault; ZF, Zarcilla fault; DGF, Don Gonzalo fault; EPF, El Puerto fault; LPF, La Paca fault; CPF, Cerro Pelado fault; 1, Triassic; 2, Jurassic; 3, Cretaceous; 4, Paleogene; 5, Pliocene; 6, Plio-Quaternary; 7, Quaternary.

Table 4 Summary of Local Fault Characteristics

Estimated Length Name Label in Figure 8 Strike (km) Regime

Avile´s fault AF N40–45 11 Left-lateral strike slip Terreras fault TF N161 13 Left-lateral strike slip Zarcilla fault ZF N48 9–11 Left-lateral strike slip Don Gonzalo fault DGF N45 8 Left-lateral strike slip El Puerto fault EPF N125 6–8 Normal, right-lateral strike slip La Paca fault LPF N90 4.5 Thrust? Cerro Pelado fault CPF N75–80 4.5 Left-lateral strike slip

Macroseismic Effects of the La Paca (2005) Coulomb static stress increased more than 0.4 bars after the Earthquake and Ground Motion 2002 event. The two major aftershocks also occurred in areas of increased Coulomb stress. In summary, stress change Damage Distribution modeling supports the fact that an earthquake-triggering pro- The 2005 La Paca earthquake caused remarkable dam- cess is active in the area. age to traditional and engineered structures, considering its 682 B. Benito et al.

Figure 9. Models of change in static Cou- lomb failure stress (CFS) generated after the events of 2002 and 2005. Dashed line represents the line of no change. (a) and (b) Changes in CFS produced on planes northeast–southwest (parallel to the fault assumed to be the source of the 2005 earthquake) by the two plane solutions of the 2002 event. (c) and (d) Changes in the CFS produced by the 2002 and 2005 mainshocks on planes optimally oriented under the current tectonic regime. All models are calculated for the hypocentral depth of the 2005 mainshock (3 km).

is shown in Figure 10. Maximum EMS intensities of VI and VII were observed at La Paca and Zarcilla de Ramos, respectively, based on macroseismic fieldwork carried out on site. Other locations with shorter epicentral distances, such as the town of Avile´s, experienced lighter intensity values (EMS V). Numerous families were housed provisionally in tents during the first days of the seismic series, while damage assessments were carried out in their homes. Emergency services from the region of Murcia including firemen, Red Cross, and civil defense personnel, supported by architects and psychologists, were deployed to aid some 2000 affected people in the macroseismic area. As a consequence of the mainshock, many buildings in the affected towns were rendered uninhabitable, most of them in Zarcilla de Ramos. More than 800 damage reports (about 60% of the buildings of La Paca and Zarcilla de Ra- mos) were claimed. More than 20 houses were initially listed Figure 10. Intensity map (EMS scale) for the as unsafe and demolished. Most of them were poor-quality La Paca earthquake (29 January 2005). traditional structures, evidencing once again the poor performance of unreinforced masonry structure. However, considerable nonstructural damage and slight structural damage was reported to code-compliant reinforced concrete frame moderate magnitude of Mw 4.8, and it caused great alarm in buildings, despite only moderate shaking. local neighborhoods. Intensity distributions show a consid- No major field effects were observed, although different erable drift from the epicentral location toward the south- locations in the epicentral area were still littered with large southwest Avile´s fault, probably associated with local ground block falls originating from the 2002 series. Preliminary es- conditions (see next section). An isoseismal map for this event timated losses of the earthquake are about 10 million €. An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005 683

Damage patterns according to building types are presented probably represents the largest risk in moderate earthquakes next. because this type of damage falls out into crowded streets with light shaking. Some examples of damage in this type of structure are shown in Figure 11b–e. Damage to Unreinforced Masonry Structures (URM). URM structures form the main building stock of pre-1960s construction in towns and cities across Spain. Masonry is Damage to Reinforced Concrete Structures (RC). The still a significant choice for small-scale, normal technology main damage trends observed in RC buildings was damage and economic construction, in general, used for housing in to nonstructural elements like walls and infill panels. There the form of concrete masonry units (CMUs). This type was ample evidence of partition walls becoming loaded dur- of masonry should be reinforced to be code-compliant ing the earthquake as shown by widespread shear cracking. (NCSE-94, NCSE-02), but there is little evidence CMUs are Excessive drift and deflection of RC structures is the main actually reinforced in ordinary construction in current cause of this type of damage, resulting in unforeseen loading practice. and participation of partition walls. The combination of rigid The most significant damage trend observed is the loss partitioning walls and RC frames are always problematic un- of connection between masonry walls due to inertial forces. less the latter are properly stiffened to limit drift and deflec- This type of failure is common in fieldstone construction tion, or the former are strengthened to properly participate because true bonds between walls are difficult to achieve as shear walls (Fig. 12a). In any case, the need for improve- due to the nature of the material. Large inertial forces are ment in the seismic performance of RC frames with brick generated perpendicular to the plane of the walls as the infill panels is clear, and this should form the basis for spe- ground shifts during an earthquake, causing them to crack cific and unambiguous attention in further revisions of the and drift apart from each other. Figure 11a illustrates the Spanish Building Code. Some examples of the damage ob- deformation and rupture process. Damage of this type is typ- served in this type of structure are shown in Figure 12b–d. ically grade 2 in buildings of vulnerability A or B. Advanced failure results in out-of-plane wall collapse and is typically Objects and Furniture. In many households furniture was grade 3 or 4. Unsupported gable walls are particularly prone displaced about 10 cm from their original locations and ob- to this type of damage. Damage to nonstructural elements jects fell from shelves, covering the floors of supermarkets such as cornices, eaves, and chimneys was widespread and with fallen produce. In some households cabinet or refrig-

Figure 11. Masonry buildings: (a) Model for out-of-plane wall drift; (b)–(e) Ex- amples of observed damage to these types of buildings. 684 B. Benito et al.

Figure 12. Concrete frame buildings: (a) Deformation process and shear transfer to nonstructural walls through lack of rigidity; (b, (c), and (d) Examples of damage observed in one of these buildings; (e) Objects thrown down from shelves in the same location. erator doors were flung open and their contents thrown out. According to the NEHRP classification, the amplifica- Many objects were thrown down from shelves (Fig. 12e), tion effect in the earthquake area (Table 5) varies from nil including large objects such as TV sets. (classes A and B), where bedrock is exposed (hard to medium rocks represented by dolomite and limestone and of Jurassic and Cretaceous age), to high (class D) where Plio- Geotechnical Studies Quaternary fluvial sediments are present. Some igneous

The aforementioned structural damages of the La Paca rocks (class C1 and C2) are also found in the area. Because earthquake in La Paca and Zarcilla de Ramos seem to depend of the similarity and the heterogeneous nature of Quaternary on not only the magnitude of the earthquake and the distance sediments, in the surface geology map, it was not possible from the source, but also on local geological conditions that to identify soil class E in this area. could have resulted in an increase in the intensity of ground The greater part of the soil (which in the geological map motion in both towns. In this regard, we have carried out a corresponds to the Plio-Quaternary fluvial and alluvial fans) detailed study of the geological materials in the epicentral in the epicentral area, belongs to either the stiff C2 silty-sand area to determine their possible contribution to local ground soils, interspaced with some cemented conglomerates, or to amplification. the soft to medium stiff silty-sand soil, corresponding to A general engineering classification of the geological class D. Zarcilla de Ramos and La Paca are located mainly materials based on the existing 1:50,000 scale surface ge- in this soil class, which exhibits high amplification according ology map of the area from IGME (Baena Perez, 1972; to the NEHRP classification. Avile´s, on the other hand, with Kampschuur et al., 1972a) was done to evaluate possible minor damage and nearest to the epicenter, is located in soils earthquake amplifications for each geological unit. The of C2–D classes (high to moderate amplification). methods proposed by Borcherdt (1994a, b) and contained on In addition, a detailed evaluation of soil amplification the 2003 National Earthquake Hazards Reduction Program factors was carried out based on the analysis of samples (NEHRP) Provisions (Building Seismic Safety Council, collected on site in the three towns. This was done mainly 2003) were adopted for this purpose. to confirm the previous surface geology classification. The An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005 685

Table 5 Categories for Soil Susceptibility to Ampliﬁcation in the Epicentral Area of the 2005 La Paca Earthquake, Based on Data from the Geological Map

Estimated Susceptibility to

Soil Type Engineering Classiﬁcation Description on the Geological Map Vs (m/sec) Ampliﬁcation Category A Hard rocks Jurassic limestone, dolomites and Ophites 1500 Nil Category B Medium strength rocks Jurassic and Cretaceous interbedded limestones 750–1500 Low and marly clays

Category C1 Soft rocks and hard soils (very stiff plastic clays) Kuiper, varied color plastic clays and gypsum 450–750 Moderate Category C2 Stiff cohesive soils; interbedded hard cohesionless Tertiary and Plio-Quaternary fluvial and 350–450 Moderate-high and cohesive soils. alluvial calcareous cemented conglomerates Category D Stiff medium soils; small total thickness of soft to Plio-Quaternary fluvial and alluvial loose silty 350–180 High medium soils. sediments Category E Soft soils with total thickness 10 m. These soils are not identified very well in the 180 Very high geological map, but occur frequently in the area. A detailed study to identify them is necessary.

Definitions are based on the soil classification of the 2003 NEHRP Provisions. results confirm that the soils in the epicentral area that sification map for the epicentral region, shown in Figure 13. underlay the buildings are mainly unconsolidated and non- A major change regarding the original classification is intro- cemented-detrital, colluvial, and alluvial sediments of con- duced in Zarcilla de Ramos, where a class D has been tinental origin. However, when we compare the results of changed to class E (soft clays). In this sense, it can be con- their geotechnical properties, differences between the soils cluded that the soils underlying the damaged areas, espe- underlying the three localities (Avile´s, La Paca, and Zarcilla cially in Zarcilla de Ramos, may have played an important de Ramos) are found. role in seismic amplification. In Zarcilla de Ramos the soils are mainly soft soils 10 m thick or more, consisting of a well to medium graded loose Ground Motion sandy-silt and gravel, interbedded with a semirigid silty clay Several strong-motion stations from the IGN network layer. Plasticity is low and the bulk dry density for these recorded the 29 January 2005 La Paca earthquake (see Table layers also is relatively high compared with the surface soils. 6). Most of these stations were located at epicentral distances The natural water content for the whole sediments is less farther than 20 km and, consequently, the recorded PGA val- than their plastic limit, which suggests that the stiffness of ues were relatively low (less than 0.024g). The distribution the interbedded silty clay layer is the result of an apparent of recorded ground motions is illustrated in Figure 14a. Note cohesion by suction. During wetting and earthquake shak- the larger amplitude recorded in the stations located in the ing, these soils can undergo an important reduction in their northeast–southwest direction compared with the ones lo- consistency, behaving as soft soils. Although the soil at cated to the southeast with respect to the epicenter. A com- Zarcilla de Ramos was classified as NEHRP type D, results bination of local characteristics at the recording sites and a of our geotechnical study show that a NEHRP type E, with channeling effect of seismic energy along the northeast– a high-amplification effect, is more appropriate. southwest direction, parallel to the main tectonic faults in Soils in La Paca are more homogenous than in Zarcilla the area, may explain this observation (Gaspar-Escribano and mainly consist of well-sorted or badly graded silty sands. and Benito, 2007). Their thickness ranges approximately from 3 to 5 m. The To estimate the impact of ground motion in the dam- water content is near to the plastic limit and their bulk dry aged towns of the epicentral areas, we followed the approach density is relatively high in comparison with the Zarcilla de of Gaspar-Escribano and Benito (2007) and applied it to ס ס Ramos soils. Thus, a NEHRP class D is confirmed for this Zarcilla de Ramos (Rep 10 km) and La Paca (Rep 5 km). soil. The simulation method of Sabetta and Pugliese (1996) Soils in Avile´s consist of a superficial (1–2 m) reddish is used to reproduce the acceleration time-histories for these brown silty-clay with some gravels and a thick (more than towns, taking Mw 4.8 and the soil factors established in the 30 m) calcareous-cemented Plio-Quaternary conglomerates previous geotechnical study (Fig. 14b). Predicted PGA val- and marls (very hard and stiff soils), which are more con- ues are about one order of magnitude larger than the re- sistent than the soils in the Zarcilla de Ramos and La Paca. corded PGAs at the stations. At the same time, the empirical A class D–C2 is thus assigned, in agreement with the first ground-motion models of Ambraseys et al. (1996), Sabetta geological classification. and Pugliese (1996), and Berge-Thierry et al. (2003) are also Taking into account the geological and geotechnical used to estimate response spectra at Zarcilla de Ramos and data in the selected areas, we propose a geotechnical clas- La Paca. The magnitude and distance definitions and specific 686 B. Benito et al.

Figure 13. Map of soils in the La Paca earthquake epicentral area, indicating the seismic-ampliﬁcation susceptibility categories. The pictures show representative soils of the area (see locations on the map).

Table 6 Location of IGN Strong-Motion Stations and Recorded PGA Values of the 29 January 2005 Earthquake

Longitude Latitude Repic PGA NS PGA V PGA EW Station Code ( W) ( N) (km) (g) (g) (g)

Mula MUL 1.49 38.04 27 0.024 0.002 0.024 Lorca LOR 1.70 37.68 28 0.007 0.004 0.006 Alhama de Murcia AHM 1.43 37.85 32 0.006 0.004 0.003 0.011 0.008 ן Ve´lez Rubio VLR 2.07 37.65 42 Murcia M02 1.13 37.98 56 0.003 0.003 0.003 Jumilla JUM 1.33 38.48 72 0.006 0.004 0.005 Olula del Rı´o OLU 2.46 37.35 80 0.002 0.001 0.001 Jae´n JAE 3.79 37.77 179 0.002 0.001 0.002

parameters used for each site, after the necessary adjust- For the site at La Paca, the NCSE-02 response spectrum ex- ments, are listed in Table 7, together with the PGA values ceeds the calculated spectra for periods longer than 0.15 sec. estimated using the different methods cited earlier. Average For the site at Zarcilla de Ramos, the NCSE-02 response PGA values around 0.1g can be derived for both sites. spectrum exceeds the calculated spectra for the whole-period The empirical and simulated spectra are plotted in range. The damage pattern observed in Zarcilla de Ramos, Figure 14c, together with the response spectrum derived more severe than in La Paca with similar vulnerability of from the Spanish Building code NCSE-02 for a return period buildings, suggests that the ground motion may have been of 475 years. All the estimated spectra for both sites consis- stronger than the one predicted by simulations and empirical tently show that maximum spectral accelerations (SA) are models. reached in intermediate periods (from 0.1 through 0.3 sec), The underestimation of the spectral accelerations may although their amplitudes vary depending on the model used. be the result of the soil factor used, which does not reﬂect An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005 687

Figure 14. (Top) Geographical distribution of recorded ground motions (north–south components are shown, except at station VLR, where the east–west component is shown). The star indicates the location of the epicenter and the triangles the location of the IGN stations. (Middle) Simulated acceleration time histories using the method of Sabetta and Pugliese (1996) at Zarcilla de Ramos (left) and La Paca (right). (Bottom) NCSE-02 and predicted response spectra at Zarcilla de Ramos (left) and La Paca (right).

accurately the elevated soil ampliﬁcations that apparently pected. During one of these events, it appears that the NCSE- occurred in Zarcilla de Ramos. This is clear evidence for the 02 response spectra would be exceeded to an even larger need to develop detailed geotechnical studies for providing degree than that observed in the present study, questioning transfer functions for these soil types. More realistic re- the spectra proposed in the code. sponse spectra would be more in accord with the ground motions experienced, helping to explain the observed Summary and Conclusions damage. The need to deﬁne more accurately local effects in This article presents a complete overview of the 29 ground motion should be extended to the response spectra January 2005 La Paca earthquake. The earthquake sequence included in the Spanish Building Code NCSE-02. This is is composed of two subevents of magnitudes Mw 4.8 and 4.2. particularly important for the area of study because earth- Moment-tensor inversions show a predominately strike-slip quakes of larger magnitude than those analyzed are ex- movement with a minor normal component for the main- 688 B. Benito et al.

Table 7 Parameters Used in the Different Models for Estimating PGA and Spectral Accelerations in the Epicentral Area (Sites La Paca and Zarcilla de Ramos)

Parameters Used in the Different Models La Paca Zarcilla de Ramos

Type of distance (km) Epicentral (Rep) 5.0 10.0 Joyner–Boore (Rjb) 1.5 5.0 Hypocentral (Rhyp) 6.5 11.0 Ground-motion model, soil Ambraseys et al. (1996) Stiff soil (0.23g) Soft soil (0.15g) type considered (and Sabetta and Pugliese (1996) Shallow alluvium (0.11g) Deep alluvium (0.11g) estimated PGA) Berge-Thierry et al. (2003) Soil (0.17g) Soil (0.10g) Simulation Shallow alluvium (0.12g) Deep alluvium (0.06g) NCSE-02 Soil III Soil IV

Earthquake magnitude Mw 4.7; Ms 4.7; mbLg 4.7 PGA values (in g) estimated in both locations, using different empirical ground-motion models and a simulation method, are also indicated.

shock, and a larger normal component for the main after- Acknowledgments shock. These results are consistent with a double-rupture The Spanish Instituto Geogra´fico Nacional (IGN) is acknowledged for event and reveal complex source geometry, with an off-fault providing the strong-motion data used in this work. The focal mechanism origin for the aftershocks. analysis was based on high-quality seismic broadband data from Instituto Several faults with neotectonic activity are identified in Andaluz de Geofisica, IRIS, and ROA/UCM/Geofon stations, as well as the the epicentral area. The so-called Avile´s fault is the most temporary TEDESE experiment for events in 2002. The authors thank likely source of the mainshock. A simplified model of fault Associate Editor Dr. Lorraine Wolf and two anonymous reviewers for their valuable comments and corrections, which highly improved the quality and kinematics in the epicentral area is consistent with a strong comprehensibility of the article. This work has been partly financed by compressive deformation and stress concentration around projects CGL2005-04541-C03-01/BTE and CGL2005-07456-C03-03/BTE the source area of the 2005 La Paca earthquake. In addition of the Spanish Ministerio de Educacioń y Ciencia. to this stress concentration, the increase of static Coulomb stresses produced after the 2002 southwest Bullas earthquake in an adjacent area may have acted as a triggering References mechanism for the 2005 event. This result reveals the sig- nificance that triggering interactions may have in southeast- Ambraseys, N. N., K. A. Simpson, and J. J. Bommer (1996). Prediction of ern Spain and constitutes an important issue to be considered horizontal response spectra in Europe, Earthquake Eng. Struct. Dyn. in future probabilistic seismic-hazard studies. 25, 371–400. Andrieux, J., J.-M. Fontbote, and M. Mattauer (1971). Sur un modele ex- Based on a broad survey of damage in different loca- plicatif de l’arc de Gibraltar, Earth Planet. Sci. Lett. 12, 191–198. tions of the epicentral area, a maximum EMS intensity of VII Badal, J., E. Samardjieva, and G. Payo (2000). Moment magnitudes for is assigned to Zarcilla de Ramos. The most significant dam- early instrumental Iberian earthquakes, Bull. Seism. Soc. Am. 90, age trend observed in URM structures was the loss of con- 1161–1173. nection between masonry walls due to inertial forces. Dam- Baena Perez, J. (1972). Mapa y memoria explicativa de la Hoja 931 (Zar- cilla de Ramos), scale 1:50,000, IGME. age to nonstructural elements was widespread and probably Berge-Thierry, C., F. Cotton, O. Scotti, D.-A. Griot-Pommera, and represents the largest risk in moderate earthquakes. The main Y. Fukushima (2003). New empirical response spectral attenuation damage trends observed in RC buildings were damage to laws for moderate European earthquakes, J. Earthquake Eng. 7, no. 2, nonstructural elements like infill panels, revealing the poor 193–222. interaction between structural and nonstructural elements in Borcherdt, R. D. (1994a). Estimates of site-dependent response spectra for design (methodology and justification), Earthquake Spectra 10, 617– this type of construction. The need for improvement in the 653. seismic performance of RC frames with brick infill panels is Borcherdt, R. D. (1994b). Simplified site classes and empirical amplifica- evident and should be the focus of further revisions of the tion factors for site-dependent code provisions, in Proceedings of the Spanish Building Code. NCEER/SEAOC/BSSC Workshop on Site Response during Earth- Besides the high vulnerability of buildings, the severity quakes and Seismic Code Provisions, University of Southern Califor- nia, Los Angeles, November 18–20, 1992, G. M. Martin (Editor), of ground motions in the epicentral area played a significant National Center for Earthquake Research Special Publication role on the distribution of damage. Soft soil layers under- NCEER-94-SP01, Buffalo, New York. lying La Paca and Zarcilla de Ramos amplified ground mo- Braunmiller, J., U. Kradolfer, M. Baer, and D. Giardini (2002). Regional tions to a significant degree. Modeling results show that for moment tensor determination in the European-Mediterranean area— short periods predicted spectral accelerations may have ex- initial results, Tectonophysics 356, 5–22. Buforn, E., and C. Sanz de Galdeano (2001). Focal mechanism of Mula ceeded the ground-motion levels of the NCSE-02 code. (Murcia, Spain) earthquake of February 2, 1999, J. Seism. 5, 277– Therefore, the revision of the soil factors proposed in the 280. code would be useful. Buforn, E., B. Benito, C. Sanz de Galdeano, C. del Fresno, D. Munõz, and An Overview of the Damaging and Low Magnitude Mw 4.8 La Paca Earthquake on 29 January 2005 689

I. Rodrı´guez (2005). Study of the damaging earthquake of 1911, 1999 Martıńez-Dıáz, J. J., A. Rigo, L. Louis, R. Capote, J. L. Hernandez-Henrile, and 2002 in the Murcia (SE Spain) region, Bull. Seism. Soc. Am. 95, E. Carrenõ, and M. Tsige (2002). Caracterizacioń geolo´gica y sis- -me (4.8 ס motectońica del terremoto de Mula (Febrero 1999, Mb .567–549 Building Seismic Safety Council (BSSC) (2003). NEHRP Recommended diante la utilizacioń de datos geolo´gicos, sismolo´gicos y de interfer- Provisions for Seismic Regulations for New Buildings and Other ometrıá de RADAR (INSAR), Geogaceta 31, 157–160. Structures. Part 1, Provisions (FEMA 450), Building Seismic Safety Martıńez Solares, J. M., and J. Me´zcua (2002). Cata´logo sı´smico de la Council of the National Institute of Building Sciences, Washington, Penıńsula Ibe´rica (880 a. C.-1900), Monografıá 18, Instituto Geo- D.C., 340 pp. gra´fico Nacional, Madrid, 253 pp. Chinnery, M. A. (1963). The stress changes that accompany strike-slip Me´zcua, J., and J. M. Martıńez Solares (1983). Sismicidad del a´rea Iber- faulting, Bull. Seism. Soc. Am. 53, 921–932. omogrebi, Instituto Geogra´fico Nacional, Madrid, Pub. 203, 301 pp. European Macroseismic Scale 1998 (EMS-98). Cahiers du Centre Europ. Norma de la Construccioń Sismorresistente Espanõla (NCSE-94) (1994). de Geódyn. et de Seísmologie 15, 1–99. Real Decreto 2543/1994, de 29 de diciembre, por el que se aprueba Freed, A. M., and J. Lin (2001). Delayed triggering of the 1999 Hector la norma de construccioń sismorresistente: parte general y edificacioń, Mine earthquake by viscoelastic stress transfer, Nature 411, 180–183. Boletıń Oficial del Estado 33, 3935–3980. Galbis, J. V. (1932). Cata´logo sı´smico de la zona comprendidad entre los Norma de la Construccioń Sismorresistente Espanõla (NCSE-02) (2002). meridianos 5Ey20W de Greenwich y los paralelos 45Ny25N, Real Decreto 997/2002, de 27 de septiembre, por el que se aprueba Tomo I, Instituto Geogra´fico, Catastral y de Estadı´stica, Madrid, 567– la norma de construccioń sismorresistente: parte general y edificacioń 569. (NCSR-02), Boletıń Oficial del Estado 244, 35,898–35,967. Gaspar-Escribano, J. M., and B. Benito (2007). Ground-motion character- Okada, Y. (1992). Internal deformation due to shear and tensile faults in a ization of low-to-moderate seismicity zones and implications for seis- half space, Bull. Seism. Soc. Am. 82, 1018–1040. mic design: lessons from recent, Mw 4.8, damaging earthquakes in Omori, F. (1884). On the aftershocks of earthquakes, J. Coll. Sci. Imperial southeast Spain, Bull. Seism. Soc. Am. 97, no. 2, 531–544. Univ. 7, 111–200. Geller, R. J., and C. S. Mueller (1980). Four similar earthquakes in Central Pollitz, F. F. (2002). Stress triggering of the 1999 Hector Mine earthquake California, Geophys. Res. Lett. 7, 821–824. by transient deformation following the 1992 Landers earthquake, Bull. Harris, R. A., R. W. Simpson, and P. A. Reasenberg (1995). Influence of Seism. Soc. Am. 92, 1487–1496. static stress changes on earthquake locations in southern California, Pondrelli, S., A. Morelli, G. Ekstro¨m, S. Mazza, E. Boschi, and A. M. Nature 375, 221–224. Dziewonski (2002). European-Mediterranean regional centroid- Hermes, J. J. (1978). The stratigraphy of the Subbetic and the southern moment tensors: 1997–2000, Phys. Earth Planet. Interiors 130, 71– Prebetic of the Velez Rubio-Caravaca area and its bearing on trans- 101. current faulting in the Betic Cordilleras of southern Spain, Proc. K. Rey Pastor, A. (1936). Sismicidad de las regiones litorales espanõlas del Ned. Akad. Wet. 81, 1–54. Mediterrańeo. II Regioń Be´tica y Subbe´tica, Geol. Chaines Betique Hermes, J. J. (1985). Algunos aspectos de la estructura de la Zona Sub- Subbetique 4, no. 1, 20–22. be´tica (Cordilleras Be´ticas, Espanã Meridional), Estudios geol. 41, Rey Pastor, A. (1949). Comarca sı´smica de Caravaca y el sismo del 23 de 157–176. junio de 1948, Instituto Geogra´fico y Catastral, Madrid, 21–33 (in Instituto Geogra´fico National (IGN) (1999). Serie sı´smica de Mula (Murcia), Spanish). Segundo Informe General, Subdireccioń General de Geodesia y Geo- Sabetta, F., and A. Pugliese (1996). Estimation of response spectra and fı´sica, Instituto Geogra´fico Nacional, Madrid, 35 pp. simulation of nonstationary earthquake ground motions, Bull. Seism. Jaume, S. C., and L. R. Sykes (1992). Changes in state of stress on the Soc. Am. 86, no. 2, 337–352. southern San Andreas fault resulting from the California earthquake Sanchez Lozano, R., and A. Marıń (1912). Estudio relativo a los terremotos sequence of April to June 1992, Science 258, 1325–1328. ocurridos en la provincia de Murcia en 1911, Bol. Inst. Geol. 12, 179– Jerez Mir, L., F. Jerez Mir, and G. Garcıá-Monzoń (1974). Mapa Geolo´gico 214. de Espanã. Scale 1:50,000, no. 912, Mula. Instituto Geolo´gico y Mi- Sanchez Navarro-Neumann, M. M. (1911). Los recientes terremotos mur- nero de Espanã, 30 pp., 1 fold map. cianos, Rev. Soc. Astr. Espanã 119–122. Jimenez Cisneros, D. (1911). Bol. Real Soc. Espanõla de Historia Natural Sanchez Navarro-Neumann, M. M. (1912). Enumeracioń de los terremotos 11, 210–211. sentidos en Espanã durante el anõ 1911, Bol. Real Soc. Espanõla Hist. Kampshuur, W., C. W. Langemberg, J. Baena, F. Velando, G. Garcıá- Nat. 512–518. Monzoń, J. Paquet, and H. E. Rondeel (1972a). Mapa geolo´gico Sanchez Navarro-Neumann, M. M. (1917). Ensayo sobre la sismicidad del Nacional a escala 1:50,000. n 932, Coy. Instituto Geolo´gico y Minero suelo espanõl, Bol. Real Soc. Espanõla Hist. Nat. 100–101. de Espanã 1 fold map Mapa y memoria explicativa de la. Sanchez Navarro-Neumann, M. M. (1920). Bosquejo sı´smico de la Pen- Kampshuur, W., C. W. Langemberg, H. E. Rondeel, J. A. Espejo, A. ńsulaı Ibe´rica, in La estacioń sismolo´gica y el observatorio astron- Crespo, and R. Pignatelli (1972b). Mapa y memoria explicativa de la o´mico de Cartuja, Granada, 43–45. Hoja 953 (Lorca) del Mapa geolo´gico Nacional a escala 1:50,000, Sanz de Galdeano, C. (1983). Los accidentes y fracturas principales de las IGME. Cordilleras Be´ticas, Estud. Geol. 39, 157–165. King, G. C., R. S. Stein, and J. Lin (1994). Static stress changes and the Sanz de Galdeano, C. (1990). Geologic evolution of the Betic Cordilleras triggering of earthquakes, Bull. Seism. Soc. Am. 84, 935–953. in the western Mediterranean, Miocene to present, Tectonophysics, Langston, C. A., J. S. Barker, and G. B. Pavlin (1982). Point-source inver- 172, 107–119. sion techniques, Phys. Earth Planet. Interiors 30, 228–241. Sanz de Galdeano, C., and E. Buforn (2005). From strike-slip to reverse Mancilla, F., C. J. Ammon, R. B. Herrmann, and J. Morales (2002). Fault- reactivation: The Crevillente fault system and seismicity in the Bullas- ing parameters of the 1999 Mula earthquake, southeastern Spain, Mula area (Betic Cordillera, southeast Spain), Geol. Acta. 3, 241– Tectonophysics 354, 139–155. 250. Martıńez-Dıáz, J. J. (1998). Neotectońica y Tectońica Activa del sector Stein, R. S. (1999). The role of stress transfer in earthquake recurrence, centrooccidental de Murcia y Sur de Almerıá, Cordillera Be´tica Nature 402, 605–609. (Espanã), Ph.D. Thesis, Universidad Complutense de Madrid, 466 pp. Stich, D., C. J. Ammon, and J. Morales (2003). Moment tensor solutions Martıńez-Dıáz, J. J., E. Masana, J. L. Hernańdez-Enrile, and P. Santanach for small and moderate earthquakes in the Ibero-Maghreb region, (2001). Evidence for coseismic events of recurrent prehistoric defor- J. Geophys. Res. 108, 2148, doi 10.1029/2002JB002057. mation along the Alhama de Murcia fault, southeastern Spain, Geol. Stich, D., F. Mancilla, D. Baumont, and J. Morales (2005). Source analysis Acta 36, no. 3–4, 315–327. of the Mw 6.3 2004 Al Hoceima earthquake (Morocco) using regional 690 B. Benito et al.

apparent source time functions, J. Geophys. Res. 110, B06306, doi ETSI Topografıá, Geodesia y Cartografıá 10.1029/2004JB003366. Universidad Politećnica de Madrid Stich, D., E. Serpelloni, F. Mancilla, and J. Morales (2006). Kinematics of 28031 Madrid, Spain (B.B., J.M.G.-E., M.J.G.R., M.E.J.) the Iberia-Maghreb plate contact from seismic moment tensors and GPS observations, Tectonophysics 426, 295–317. Department de Geodina´mica, F. CC. Geolo´gicas Toda, S., and R. S. Stein (2003). Toggling of seismicity by the 1997 Ka- Universidad Complutense de Madrid goshima earthquake couplet: A demonstration of time-dependent 28003 Madrid, Spain stress transfer, J. Geophys. Res. 108, B12, 2567, doi 10.1029/ (R.C., J.J.M.-D., M.T., J.M.I.-A., J.A.A.-G., C.C.) 2003JB002527. Toda, S., R. S. Stein, P. A. Reasenberg, and J. H. Dieterich (1998). Stress Architect Kobe, Japan, shock: Effect on after- Piamonte 18, 4E 28004 Madrid, Spain 6.5 ס transferred by the Mw shocks and future earthquake probabilities, J. Geophys. Res. 103, (P.M.) 24,543–24,565. Velando, F., and J. Paquet (1974). Mapa Geolo´gico de Espanã, Scale Istituto Nazionale di Geofisica e Vulcanologia 1:50,000, no. 911. Cehegıń, Instituto Geolo´gico y Minero de Espanã, 40128 Bologna, Italy 28 pp., 1 fold map. (D.S.) Wells, D. L., and K. J. Coppersmith (1994). New empirical relations among magnitude, rupture length, rupture width, rupture area and surface Laboratorio de Geotecnia (CEDEX) displacement, Bull. Seism. Soc. Am. 85, 1–16. Alfonso XII, 3 Madrid 28014, Spain Zeng, Y. (2001). Viscoelastic stress-triggering of the 1999 Hector Mine (J.G.-M.) earthquake by the 1992 Landers earthquake, Geophys. Res. Lett. 28, 3007–3010. Manuscript received 28 July 2005.